help creating a shaded material

[bubs]

HA! I just did the same. There is only a small difference, it's hard to see, but if you flick between these 2 you can see that the image with gamma 1.0 has more chrome material... but the point of transition seems to be centred, which it probably should be with this map.

If you check the 'smooth' however, the gamma change seems to make no difference

Either way, I think the material is pretty close to the reference image!

[galinette]

You are both right and wrong...

When you do a linear black>white gradient in PS, it creates RGB values linearly mapped between 0 and 255. Whatever the gamma setting in PS is (in fact, PS, as most image software, work in gamma space without knowing anything about gamma), if your gradient is 256 pixels wide, the RGB value will increment by 1 at each pixel.

This means that the gradient is non linear in luminance space (ie physical space, linear space) but since the gamma=2.2 is somewhat matching the non linear nature of human vision, you feel that the gradient is linear... A photoshop mistake is roughly compensated by the human vision at the end.

Exception : if photoshop is in 32bit/channel mode, then it works in linear space. Just try making a gradient in 32bit mode : it will produce different results than creating a gradient in 8bit and converting to 32bit!

Back to Indigo : for linear gradient values between 0 and 255, if you use a gamma of 1.0, Indigo will produce linear values between 0.0 and 1.0 (simple division by 255). But if you use a gamma of 2.2, indigo will produce a non linear gradient.

Conclusion:
- Linear 8bit gradient used for blending : using gamma=1.0 will produce linear blending
- Linear 8bit gradient used for displacement : using gamma=1.0 will produce linear displacement
- Linear 8bit gradient used for color (diffuse, ...) : using gamma=2.2 will produce a perceptually linear gradient as you will generally expect (but a non linear reflectance curve)

[bubs]

*ahem*... OK... I feel a bit simple now...

[galinette]

HA! I just did the same. There is only a small difference, it's hard to see, but if you flick between these 2 you can see that the image with gamma 1.0 has more chrome material... but the point of transition seems to be centred, which it probably should be with this map.

Either way, I think the material is pretty close to the reference image!

In fact, gamma=1.0 gives the exact same amount of chrome and glass. But since reflectance of chrome is high, and of glass is low, the chrome "eats" the visual impression.

[galinette]

HA! I just did the same. There is only a small difference, it's hard to see, but if you flick between these 2 you can see that the image with gamma 1.0 has more chrome material... but the point of transition seems to be centred, which it probably should be with this map.

If you check the 'smooth' however, the gamma change seems to make no difference

Either way, I think the material is pretty close to the reference image!

Nice stuff, but almost impossible to manufacture at reasonable cost. An industrial design would use a serigraphied chrome layer, using variable size dots for the transition between glass and chrome.

[bubs]

elDiego, galinette says you are not allowed to use your material, it is too expensive!

[Oscar J]

Fascinating stuff guys.

[bubs]

In fact, gamma=1.0 gives the exact same amount of chrome and glass. But since reflectance of chrome is high, and of glass is low, the chrome "eats" the visual impression.

That's actually what I meant... I was trying to say that the gamma 1.0 had more chrome than the 2.2 version, not more chrome than glass. Sorry that wasn't very clear.

[elDiego]

Thanks guys! I'm glad that i generated this discussion.

I know that in real life it would probably be a layer of dots that fade. but good luck doing that in sketchup.
it could also just be spray mirror paint on the surface (which would look more like the indigo render.
I still think theres a white shade in between the 2 materials, I will try to play with the shade and the gamma.

did you notes it that fireflies are going crazy on my shaded material. do you think thta has to do with the gama? do you have any other times on that?

[galinette]

I know that in real life it would probably be a layer of dots that fade. but good luck doing that in sketchup.

Nothing related to sketchup! All you need is a blend texture which has some dots instead of a gradient.

Such as this one which was made for a real-life glass serigraphy:

Also, besides being more manufacturable, using a black/white serigraphy instead of a gradient would allow you to use the "force step" blend material option, which would likely answer your second question about fireflies

[galinette]

This one should be OK for you

Don't forget to activate "step" in blend material

Etienne

[Zom-B]

did you notes it that fireflies are going crazy on my shaded material. do you think that has to do with the gamma? do you have any other times on that?

I recently rendered gold graded bottle for a customer and found that MLT with a very high value of "Large mutation Prob." of 0.95 worked best! Also a SSS factor of 3 helps a lot!

Great "trick" with the doted gradient galinette, will need to check that out!

[elDiego]

Thank you again guys, i'll try with the serigraphy

[bubs]

This one should be OK for you

Great! this looks fantastic! You get that wonderful distortion of the light due to the dots.

I have to find some excuse to use this material now!

[zeitmeister]

Nice Trick with the dotmap.
Activating step blend renders faster and cleaner.
Very cool!